The Development of Hillforts in Wessex
Text by Barry Cunliffe
Before Danebury Ring was built hillforts were rare but not unknown in Wessex. In the immediate area the large defended enclosure of Balksbury was already in existence. Its rampart and ditch, while modest in size, were refurbished on a number of occasions. Evidence of use is not extensive but it is clear from the recent excavations that livestock was kept inside for periods of time. Further afield there are other forts of similar date and form: Harting Beacon (Sussex), near Petersfield; Bathampton Down, near Bath; and Bindon Hill near Lulworth (Dorset).
All were large (Bindon is massive), all had very considerable areas of open space within and all seem to have been occupied, albeit sparsely, in the period roughly 800-600 BC. There were no doubt many more: Walbury (Hampshire), and Ogbury and Martinsell (Wiltshire) share superficial characteristics with the excavated examples but have not themselves been examined. Elsewhere, at Ham Hill (Somerset) and Hod Hill (Dorset), later occupation has partially obscured the earlier features. The list, though very incomplete, is sufficient to indicate that large, early hilltop enclosures were widely, if sparsely, distributed over the Wessex landscape.
Functionally these sites are very difficult to interpret but at the very least their size implies considerable and concerted effort on the part of the community and they are likely to have fulfilled communal functions as places for social gatherings and annual fairs associated with rounding up and corralling of livestock, and as storage and collection centres – all the activities necessary to maintain the cohesion of a large rural community.
What is particularly noticeable is that, without exception, all the excavated sites were abandoned by the middle of the first millennium, though a few were reused for a variety of functions centuries later. If this observation is borne out by further work, then the demise of the early hilltop enclosures must be seen as a major threshold in the socio-economic development of Wessex. The exact meaning of the decline is obscure but it could well represent a change from a loosely integrated and rather dispersed society to a more centralised form of socio-political organisation.
The next stage is quite dramatic. Some time, probably within the sixth century BC, a rash of quite substantially built hillforts appears all over Wessex. Unlike the early hilltop enclosures these forts were smaller, averaging 5-6ha (13 acres), and rather more strongly defended. This was the stage at which Danebury Ring was built and it was at about this time that the nearby site of Bury Hill was first defended. Winklebury, near Basingstoke, also belongs to this phase. A little later, in the fifth century, new forts were built at Quarley Hill, Figsbury Hill and Woolbury all of about the same size and strength and each able to command equivalent sized territories.
The earliest phase of Danebury Ring is comparatively well known as the result of our current excavation. An area of about 5ha (12 acres) was enclosed by a single rampart fronted by a ditch and was provided with two entrances, on opposite sides of the enclosure, between which ran a roadway (34 & 35). There were probably subsidiary roads but their lines are not absolutely clear. Much of the central part of the site to the north of the main road was occupied firstly by rows of four-post granaries (or so they are generally thought to be), but these were soon replaced by a densely packed mass of storage pits. Superficially, it could be argued that this signifies a change in storage technique but the argument is weak, for the evidence need imply little more than a minor rearrangement with pits replacing granaries in one area while, perhaps, new granaries were being built in another. Without a total excavation of the site the question will have to remain unresolved.
To the south of the main road there appears to have been a totally different kind of arrangement. Here were a number of circular houses with the spaces between occupied by groups of granaries and pits. Finally, around the periphery of the settlement in the shelter of the rampart, there is some evidence to suggest the clustering of circular houses. The contrast between the north and south parts of the enclosure is particularly interesting. Perhaps what is apparent here is a division between the resident community in the southern zone, each family or family group with its own food stores, and in the northern area facilities for communal storage of surplus grain from the farmsteads outside the fort.
While much can be said about Danebury Ring in the sixth and fifth centuries, comparatively little is known of other contemporary forts; but in Dorset, at both Maiden Castle and Chalbury, sufficient work was done to show that these early forts were of comparable size and strength to Danebury and were extensively occupied by communities building circular houses and digging storage pits. However, there was little that could be said of the internal arrangements of the settlements.
Returning now to the Danebury Ring region, there is sufficient evidence to show that in the period c.600-400 BC a number of other hillforts, including Figsbury, Quarley, Bury Hill and Woolbury, were also built but it is not certain that they were all occupied in the same way as Danebury nor were they all necessarily in contemporary use. The recent excavations at Bury Hill and Woolbury and the older, more limited, work at Quarley and Figsbury show that the picture is really quite complicated (and is liable to become more so as work progresses), but on the evidence of the defensive structures it is probable that Bury Hill and Danebury were the first forts to be established, by or soon after 600 BC. Both had timber-faced ramparts.
However, at Bury Hill no evidence of internal occupation, of the kind found at Danebury, had been identified and the possibility remains that the fort was very soon abandoned. Sometime later, probably during the fifth century, Quarley, Woolbury and possibly Figsbury were defended using a dump-rampart construction without internal timbering. It is not clear whether or not these sites were occupied for any length of time but at Woolbury, where a considerable area was excavated within the defences, only a single pit of this period was found and, as we will see, by the beginning of the third century when Danebury was massively redefended, all three had been abandoned.
To interpret this scant evidence is not easy but it is tempting to see the centuries between 600 and 300 BC as a time of change and possibly conflict when the different polities of the region were attempting to establish, or legitimise, their social positions by building hillforts. There may well have been conflict, and in this context should be noted the burning of the Danebury Ring gates sometime in the fourth century. Out of it all Danebury Ring emerged dominant.
This is one possible scenario. Another, no less possible, is that the rash of hillforts built in this period served a variety of different functions within a larger polity of which Danebury remained the centre throughout. On balance, the evidence for the burning of the Danebury gates and large quantities of sling stones found behind the southern rampart tends to favour the former interpretation. Could it be that the symbolic refurbishing of the ramparts in period 2 and the construction of the hornworks at the south-western gate at this time were a mark of the eventual victory of the Danebury Ring community over their rivals? We will never know.